Datasets:

Project-AgML
/

grapevine_growth_detection

dataset_info:
  features:
    - name: image
      dtype: image
    - name: objects
      struct:
        - name: bbox
          list:
            list: float64
        - name: categories
          list:
            class_label:
              names:
                '0': '0'
    - name: split
      dtype: string
  splits:
    - name: train
      num_bytes: 120289152
      num_examples: 1212
  download_size: 120727429
  dataset_size: 120289152
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
license: cc-by-nc-sa-4.0
task_categories:
  - object-detection
size_categories:
  - 1K<n<10K

Grapevine Growth Detection

A dataset for detection of grapevine bunches. The dataset contains 1,212 images with 3,476 bounding box annotations across 1 category.

This dataset is indexed on https://project-agml.github.io/ as part of the AgML python library.

The original train/test/val split has been preserved in the split column.

Citation

@article{schieck2023comparison,
  title={Comparison of deep learning methods for grapevine growth stage recognition},
  author={Schieck, Martin and Krajsic, Philippe and Loos, Felix and Hussein, Abdulbaree and Franczyk, Bogdan and Kozierkiewicz, Adrianna and Pietranik, Marcin},
  journal={Computers and Electronics in Agriculture},
  volume={211},
  pages={107944},
  year={2023},
  publisher={Elsevier}
}

Schieck, M. (2022, October 31). Grapevine development stages (BBCH 71-79). Kaggle. https://www.kaggle.com/datasets/schieckmartin/grapesdevelopmentstages

This dataset was reformatted from its original format to match HuggingFace standards.